Technical Hydra Performance 135 Leichtbau State of the Art Thread

Jan 31, 2017
I've decided to create a thread to chronicle all the updates and developments I've put together for my 1er over the past few months. I've given much thought to improving several aspects of our driveline, and came up with several useful observations and - dare I say - new ways of doing things that I figured I would share with the community. So without further ado, here goes:

N53 head + Schrick cams

The biggest bottleneck in the N54 when it comes to making power lies in its cylinder head design. Optimized for peak torque at very low-rpm, it becomes more and more of a restriction as more air is pumped through the motor with increasing piston speeds and pressure ratios. Even porting it for all its worth isn't enough, as the exhaust port outlet is simply too small @32mm ID for a 500cc cylinder, leading to choked flow and excessive pumping losses on the exhaust stroke. The Mitsubishi 4B11 on the other hand (a good example of a modern high-performance 500cc/cyl FI design with similarly sized exhaust valves) runs the equivalent of a 36mm exhaust port, just for the sake of comparison... The intake side isn't all that much better, with a 22.x mm venturi feeding a 31.4mm intake valve, with a horrible short-side radius which really kills high(er) lift flow on the N54 head, far from ideal once Schrick cams come into the picture... So I looked at other BMW NG6 cylinders heads for inspiration. The N52 comes with a very nice CNC ported head, a la S54, but lacks provisions for DI. Now the N53 on the other hand, available in a 90PS/L variant in its ultimate form, runs bigger valves than the N54, and best of all runs the very same DI hardware we do. So I figured surely this was worth a closer look, and went out and procured one to see for myself, and what I found was very encouraging. Pros : +0.6mm intake valves, +1mm exhaust valves, +4mm exhaust ports, 28.x mm venturi on the intake ports (ideal for a 32mm valve), and a nice sharp recessed divider on the intake ports, similar to what you would see on a modern bike engine. Best of all, stock N54 intake manifold bolts right up! Cons: Coolant passages in the deck are different, exhaust manifold flange is different, meaning custom exhaust manifolds are a must, head bolt holes are 1mm too small a la N52, and custom washers for the head bolts need to be made. So it became clear to me that I would have to do whatever it takes to make this work, as the pros easily outweighed the cons in my mind. Of course having a shop complete with machinist and fabricator might have had something to do with it.. :)

Unfortunately, for reasons which escape me, I did not document this process and only have 1x photo of the (as yet unassembled) head before it went on the motor... I will have more in the near future as I am preparing two more of these heads to go out to customers in the US, one of whom will put it on a flowbench to compare with his existing ported N54 head, I am expecting ~23x cfm @ 0.45" lift @ 28" on the intake, and ~185 on the exhaust, so watch this space!


Jan 31, 2017
Just for the sake of reference, here is a dyno plot of a stock N53, with catless headers, tune, and an N54 intake manifold, notice how it makes almost 100PS/L , and how power keeps rising with rpm while maintaining a very flat torque curve... Food for thought eh? :)

fbo n53 dyno.jpg
Jan 31, 2017
Custom TT manifolds

Now with the top end of the motor out of the way, and since custom-fabrication was going to be needed anyway, my focus turned to the exhaust manifold(s) and how best to address the matter. I have always been of the opinion that , with all due respect to everybody selling and running them, that ST setups are a vastly inferior choice on the N54 due to several reasons; packaging constraints, incompatibility with the MSD8x DME without additional control boxes, and greater rotational inertia for a given flow capacity being the first to come to mind. Our cars simply were not designed for a single (regular rotation, N55 = reverse rotation) turbocharger setup, and in my experience the closer something is to having OE-like fitment to better, and more reliable the end result is likely to be @ a systems level. So after giving this much though, I decided that the OE shorty manifold design was a very good act to follow, and that I would try and recreate the same geometry using 1.25" Sch 10 304SS elbows whilst maintaining the most compact flow path possible to maintain velocity going into the turbine housing. Here is how it turned out; note the use of fusion welds wherever possible




Last edited:
Jan 31, 2017
HP800 Turbos

Now, you may be wondering what turbos are befitting of the setup described above. The answer is a new product I've been developing since Q3 2017, that has taken an inordinate amount of my time, effort, and money to make a reality. Right from the outset, I set out to create what I felt would be the ultimate stock-frame N54 twins, offering the very best construction, aero/adiabatic efficiency, and whp/psi available while keeping the overall package as close to OE-like as possible. Here is how I did it:

-The key to achieving my goals lies in the investment-cast SS turbine housing I designed boasting the highest flowing hotside of any existing stock-frame N54 TT offering as a quick visual comparison of the volute will confirm, with a nozzle area of 6.6cm^2, or ~35% more than stock, and even 10% more than S55 turbos. This is par for the course for native TD04 turbo designs, which usually have housings in the 6-7cm^2 range. Why other upgrade turbo mfgrs keep this information secret or "proprietary" is a mystery to me, maybe they just don't know the answer?
I also opened up the volute radius some 10% over stock to help turn the flow into the turbine, and added a proper cast-in separator between the turbine exducer and the wastegate port to reduce turbulence just like the stock design - something I haven't seen recreated on any other stock-frame N54 upgrade turbo so far. This smacks of poor design and a lack of attention to detail, as it costs nothing to incorporate into a new design.

What this all means at the end of the day is lower backpressure and EGT for a given boost level = higher Volumetric Efficiency & reduced pumping losses on the exhaust stroke = more whp/psi = WIN! :)

- Moving on to the turbine wheel. Right from the beginning I decided that nothing but genuine MHI 12-bladed TD04HL turbine wheels will do, as opposed to the generic chinesium 9-bladed turbine wheels of dubious design and suspect metallurgy. On paper, an increase in blade count means a decrease in the "slip ratio" and an increase in isentropic efficiency. This comes at the theoretical cost of slightly reduced flow area @ the exducer, along with a slight increase in rotational inertia. In practice neither of these disadvantages are realized, since most aftermarket TD04 turbine wheels (that I've seen at least) appear to be of inferior quality and have thicker blades than the native 12-bladed original. Putting theory aside for a moment, every single Mitsubishi Lancer Evolution (a platform we are extremely familiar with) ever produced came factory-fitted with a 12-bladed TD05H(RA) turbine wheel. Keep in mind that this is a platform that is still successfully used in competition to this day, still claiming multiple national rally/AutoX/etc championships all over the world while running the stock turbo. Case in point being my business partner's French Blue Evo VII AutoX monster, a 6-time national AutoX championship winner, which puts out just over 520whp on race gas all while running an FP 71HTA turbo, basically a stock Evo9 TD05 turbo, with an upgraded billet compressor wheel. You may be wondering how any of this is relevant to our discussion, well the TD04HL and TD05H turbine wheels come from the same design family, with the latter being just a 16% scaled-up version of the former. The resulting aerodynamic similitude bodes very well for both performance and durability...

As for the compressor side of things, I went through the trouble of designing a bespoke 7+7 bladed high-trim compressor specifically for this application, using the latest in high specific-speed compressor design techniques whilst juggling the conflicting requirements of maximizing adiabatic efficiency while maintaining a wide map width @ high pressure ratios for surge protection. I also had it anodized in blue for extra style points lol. Note that your generic chinesium 11-blade GTX clones don't take any of this into account, (assuming they are even accurate copies of original design to begin with) as all native GTX applications use surge-ported covers... Speaking of covers, I specced a native TD04 compressor cover which offers the correct diffuser radius and volute area. A TD03-style compressor cover is simply too small to properly support the compressor wheel in question, at the cost of decreased adiabatic efficiency and flow potential.

The end result of all this is a design capable of up to 74lb/min of flow before overspeed. This means that more flow is very well possible, but then adiabatic efficiency (and durability obviously) plummet. This means that a genuine 800bhp SAE should not be asking too much , provided the use of a high-enough octane fuel of course (race gas, E85, etc)... The difference with the other TT offerings out there being that this figure will be attainable at a lower boost level, with resultant benefits to EGTs, IATs, engine, and turbocharger durability.

As for construction, it goes without saying that these use full native TD04 CHRAs and are VSR balanced @ up to 150k RPM on a UK-made Turbo Technics V3 balancer, with individual balancing reports with each serialized CHRA for additional quality assurance.

In closing let me reiterate that I am not claiming to have reinvented the wheel here. After all, the turbocharger was patented in 1905, and pretty much everything done since has more or less been just detail design changes to the original idea. What I did was merely to piece together what, in my opinion were, the optimal combination of components to suit a high performance stock-frame TT N54 setup, and I am quite certain these will make the most whp/psi of any TT option out there, and quite possibly more than most ST kits as well...




Last edited:
Jan 31, 2017
SS Turbo Heatshield & Outlet

In the interests of minimizing radiant heat within the engine bay, my buddy @frontside0815 @ inspired me to design a 0.6mm thick mirror-polished (on the inside surface) stainless steel heatshield to box in as much of the hotparts as practical, and also wrapped the downpipes for good measure to further help keep underhood temps down. I made a pilot batch of 10 of these heatshields if anyone is interested, fits right on top of the OE solenoid heatshield bracket - PM for details..

Also designed my own outlet, with full flow 1.75" ID connections, with a good shallow-angle merge into a 2.5" ID, and wrapped the whole thing in gold foil for good measure. Fits very well onto the car, and will never burn up due to being very far from any major heat sources.



Last edited:
Jan 31, 2017
3" Inlets + custom airbox

It goes without saying that all the fancy turbo hardware in the world isn't going to amount to much if you can't feed your compressor with the densest air possible. I literally spent months attempting an Alpina B3S-style setup before abandoning it for a much simpler solution you will see in the pics below, which keeps the water-tank in its stock location, while providing a full 3" of flow, enough for 350-400bhp per turbo with minimal pressure drop, along with an enclosed airbox feeding the R turbo with provisions for the stock breather and heater built-in , no messy VTA setup here!




Last edited:
Jan 31, 2017
Cooling setup

And now, to turn our attention to the cooling system.. Since I wiped out a bearing and had to replace my stock oil cooler anyway, I ended up going with an OEM N55 water to oil cooler for faster warmup and less pressure drop. Of course this adds heat rejection into the coolant circuit, so I figured I would offset this somewhat with a CSF radiator and call it a day. But after studying the 1M/335is/PPK aux rad setup closely I came to the realization that the aux cooler uses EXACTLY the same heat exchanger as our stock oil cooler! Did a bit of thinking and decided it would be both easiest and best to run this in series with the N55 oil cooler, so that the coolant going into the oil cooler is a little bit cooler than that of the engine, basically trading off a bit of coolant-side effectiveness for oil-side capacity. I can always have a custom, higher-capacity heat exchanger made in the future, but preliminary testing has shown that oil temps are noticeably cooler with this setup already...



Last edited:
  • 1Like
Reactions: Bimmer TV
Jan 31, 2017
Sooo... You didn't opt for rods, pistons, upgraded bearings, ARP hardware, or even crank hub solution? Or do you plan to during your build?
No, just a VTT bolt capture kit, and hub fix. Trouble is the hub fix was only made available literally 4 days after I spun my hub the first time I hit the limiter @ 8k. :(

My reasons for not building the motor are twofold, first of all I don't have the traction available to be able to run high boost to really make a built motor worthwhile, despite having all the "right" chassis mods. I changed out the bearings, opened up the rod and main clearances a little, and replaced all the TTY bolts, but otherwise left everything factory. I thought long and hard about adding Carrillo rods, on account them being ~50g lighter each, to make life easier on the rod bearings, but in the end decided not to sink any more money into the N54 bottom end, in the hope that sometime in the near future I may be able to swap an S55 shortblock instead, custom DME work permitting ;)


Brigadier General
Staff member
Aug 11, 2017
First off, Omar, this is a great thread. I'll be watching and my hat's off to you.

My reasons for not building the motor are twofold, first of all I don't have the traction available to be able to run high boost to really make a built motor worthwhile, despite having all the "right" chassis mods.
High boost aside, I don't know what the car's application is. We've found that can have a significant impact on decision making and problems to solve. For example I think you show a hillclimb photo are your avatar. We are looking at road course where cooling is a problem. Cooling and air management is a HUGE problem. In particular the 1 series (at one of the 1M's we are working with) has some serious deficits in that area. I digress.

One reason to consider a bottom end could be to have more low end torque on tap for exiting turns on a road course, IF the car can provide the traction to launch the cars out of the turn. This is less a high boost scenario and more of a rate of torque change. I also suspect this topic is a sister to the crank hub spinning - thinking of it as breakaway torque versus constant torque.

Just thinking out loud. Look forward to seeing this thread progress!

Last edited:

Similar threads